Tandem Mass Spectrometry

Analytical Biochemistry 288, 44 –51 (2001) doi:10.1006/abio.2000.4877, available online at http://www.idealibrary.com on

Quantification of Plasma Membrane Ergosterol of Saccharomyces cerevisiae by Direct-Injection Atmospheric Pressure Chemical Ionization/Tandem Mass Spectrometry T. H. Toh,* B. A. Prior,* and M. J. van der Merwe† ,1 *Department of Microbiology and †Department of Biochemistry, University of Stellenbosch, Matieland 7602, South Africa

Received April 28, 2000

A method for the quantification of ergosterol by atmospheric pressure chemical ionization (APcI) mass spectrometry with direct injection is described. Ergosterol and squalene were ionizable with methanol as the carrier solvent. Using positive-mode tandem mass spectrometry (MS/MS), ergosterol could be identified unambiguously without interference from structurally related compounds such as lanosterol, cholesterol, and squalene. Molecular ions of ergosterol, lanosterol, and cholesterol were detected as the [M ⴙ H ⴚ H 2O] ⴙ ion species, while squalene appeared as the [M ⴙ H] ⴙ ion species. Upon fragmentation of the three sterols and squalene, the product ion at m/z 69 was present as one of the major fragments in all four compounds. This product ion was used for the quantification of ergosterol in multiple-reaction-monitoring acquisition mode. The relationship between signal intensity and ergosterol concentration was linear over the concentration range of 0.15 to 5 ␮g/ml, or 7.56 –252 pmol ergosterol per 20 ␮l injection. The plasma membrane ergosterol of the yeast Saccharomyces cerevisiae could be quantified reproducibly without the need for prior separation from other lipids or derivatization. Six repeated injections of ergosterol standards at concentrations of 0.95 and 4.25 ␮g/ml gave standard deviations of 0.031 and 0.084, respectively, and coefficients of variation of 3.33 and 1.98%, respectively. The coefficient of variation for the four independently extracted membrane ergosterol samples was 11.18%. The presence of other lipids in a crude lipid extract did not interfere with the ergosterol determination. Direct injection APcI with multiple reaction monitoring is a

1 To whom correspondence should be addressed at Department of Biochemistry, University of Stellenbosch, Private Bag X1, Matieland 7602, South Africa. Fax: ⫹27-21-808 5863. E-mail: mjvdm@ land.sun.ac.za.

44

convenient and sensitive method for ergosterol quantification requiring no prior fractionation. © 2001 Academic Press

Key Words: atmospheric pressure chemical ionization; plasma membrane ergosterol; squalene; tandem mass spectrometry; multiple reaction monitoring.

Ergosterol is the main sterol present in the plasma membranes of Saccharomyces cerevisiae (1–3). Due to intense interest in this sterol, several methods are available for the determination of its concentration, e.g., digitonin precipitation/UV spectroscopy (4), highperformance liquid chromatography separation and detection (5, 6), and gas chromatography. Limitations caused by incomplete resolution of the different molecular species in complex mixtures can be overcome by coupling chromatographic separation to mass spectrometry. The use of GC/MS for the quantification of ergosterol has been shown to be very sensitive (7, 8). GC/MS, however, requires pretreatment of the sample prior to analysis, such as the need for chemical derivatization for ionization and detection. For compounds with molecular masses of less than 1000, the commonly used ionization techniques are electron impact ionization (EI), 2 atmospheric pressure chemical ionization (APcI), and electrospray ionization (ESI) (9 –12). EI requires introduction of the analyte in the gas phase, and the molecular ion is either small or absent due to extensive fragmentation. ESI, while compatible with liquid chromatography (LC) systems, is better suited to polar compounds. APcI was the method selected for this work and is a gentle LC-compatible ionization technique that does not produce extensive fragmenta2 Abbreviations used: EI, electron impact ionization; APcI, atmospheric pressure chemical ionization; ESI, electrospray ionization.

0003-2697/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

ERGOSTEROL QUANTIFICATION BY DIRECT-INJECTION MASS SPECTROMETRY

tion of the analyte molecule (10). The low fragmentation produced by APcI can be exploited by coupling with a triple-quadrupole mass spectrometer (9). This enables the selective fragmentation of the precursor ion, since the first quadrupole can be set to transmit a selected mass for fragmentation, the products of which are screened by the second analyzer, offering resolution and sensitivity for the study of ion fragmentations (13). Recently, several publications have appeared reporting the application of APcI/MS linked either to GC or HPLC for the quantification of neutral lipid species (14 –16). The aim of this paper is to demonstrate that the properties of APcI/MS/MS offers a means of quantifying plasma membrane ergosterol from S. cerevisiae by direct sample injection without the need for prior separation by GC or HPLC. Since preparations of S. cerevisiae may contain structurally related biosynthetic precursors of ergosterol, such as lanosterol and squalene, the method must demonstrate specificity for ergosterol in addition to sensitivity. To this end, the structural analogues lanosterol and cholesterol were included to test the specificity and robustness of the method. MATERIALS AND METHODS

Chemicals. Ergosterol (M r 396.7), lanosterol (M r 426.7), cholesterol (M r 386.66), and squalene (M r 410.7) were commercial preparations of the highest purity from Sigma–Aldrich Chemie GmbH (Steinheim, Germany). Hexane was from Riedel-de Hae¨n (Seelze, Germany), and methanol was from Romil Super Purity Solvent (Cambridge, UK). Strain and cultivation conditions. The strain used in this study was S. cerevisiae W303-1A (Mata leu2-3/ 112 ura3-1 trp1-1 his3-11/15 ade2-1 can1-100 GAL SUC2 mal0). The yeast was cultivated on YPD [2% glucose, 2% peptone (Difco), and 1% yeast extract (Difco), pH 6] medium and incubated on a rotary shaker at 30°C to late log phase. Plasma membrane isolation, ergosterol extraction, and sample preparation. Cell disruption, plasma membrane fractionation, and purity checks were carried out according to Serrano (17). Purified plasma membrane samples containing squalene as the internal standard were saponified in an alcoholic KOH solution (18) at 85°C for 90 min. Free ergosterol was extracted into hexane, which was then evaporated under a stream of nitrogen. The residue was redissolved in 2 ml ethanol, filtered, and diluted 1/20 in methanol prior to analysis. The ergosterol standards, containing squalene as the internal standard, were also saponified and extracted as described above to enable a direct comparison between standard and unknown prepara-

45

tions. Protein concentration was determined using a Bio-Rad protein assay kit (Hercules, CA). Mass spectrometry. Samples were injected in 20-␮l aliquots using a Waters 717 autosampler into a carrier stream of methanol delivered by a Pharmacia LKB 2249 (Uppsala, Sweden) gradient pump at a flow rate of 100 ␮l/min. The carrier was delivered directly to the APcI source of a Micromass (Manchester, UK) Quattro triple-quadrupole mass spectrometer. Ionization was carried out at a corona voltage of 3.5 kV, and at source and probe tip temperatures of 120 and 200°C, respectively. The cone voltage was 26 V. Multiple reaction monitoring was employed to quantify ergosterol and to detect the internal standard squalene. The precursor ions of ergosterol ([M ⫹ H ⫺ H 2O] ⫹) at m/z 380 and squalene ([M ⫹ H] ⫹) at m/z 412 were selected by the first analyzer and fragmented in the collision cell by introducing argon at a pressure of 1.8 ⫻ 10 ⫺3 mbar and applying a collision energy of 30 eV. The second analyzer was set to detect the main product ion at m/z 69 for both ergosterol and squalene. The dwell time was set at 0.2 s. A standard curve was constructed from a known concentration range of ergosterol standards by plotting the relative signal strength of m/z 69 from ergosterol/squalene against ergosterol concentration. The concentration of ergosterol in the samples was determined by comparing the ratios of the m/z 69 peak areas from ergosterol and squalene with those of the standard curve. This was accomplished by the Micromass MassLynx Quantify program. Mass spectra of the precursor ions of ergosterol, lanosterol, cholesterol, and squalene were produced by direct injection and scanning the first analyzer from m/z 200–600 at 2 s/scan. Data were acquired in the continuum mode and representative scans were produced by averaging scans across the elution peak. Product ion scans of the above compounds were obtained by fragmenting the respective precursor ions in the collision cell as described above. The product ions of each were then detected by scanning the second analyzer from m/z 45–500 at 2 s/scan. RESULTS AND DISCUSSION

In the method development for the determination of ergosterol by APcI/MS/MS, the criteria to be met were set out according to the published recommendations for method validation (19, 20): (1) Linear correlation of signal intensity with ergosterol concentration over the concentration range of interest; (2) sufficient sensitivity and precision for the determination of membrane ergosterol content; (3) demonstration of specificity and robustness such that the detection process was not susceptible to interference by structurally related compounds. Thus, two structurally related sterols lanosterol and cholesterol were also tested.

46

TOH, PRIOR, AND VAN DER MERWE

FIG. 1. APcI mass spectra showing the parent ions and the corresponding m/z of (A) ergosterol, (B) lanosterol, (C) cholesterol, and (D) squalene.

Mass Spectra Ionization of ergosterol, lanosterol, and cholesterol by APcI in the positive mode, using methanol as the carrier solvent, produced the [M ⫹ H ⫺ H 2O] ⫹ ion with m/z corresponding to 380, 410, and 370, respectively (Figs. 1A–1C). The normal mechanism of cationization for APcI is protonation at a basic site on the molecule. In the case of sterols, the loss of the secondary alcohol by dehydration has been proposed for cholestane (21).

The susceptibility for dehydration appears to be dependent on the ring conformation and, hence, the proximity of adjacent hydrogen atoms. The loss of water from ring A of ergosterol is also enhanced by the ⌬ 5 or especially the ⌬ 5,7 unsaturation (22). Squalene was ionized by protonation, yielding an [M ⫹ H] ⫹ ion at m/z 412 (Fig. 1D). Under the APcI conditions employed here, only the dehydration products of the protonated molecules were observed for the sterols. Ionization was

ERGOSTEROL QUANTIFICATION BY DIRECT-INJECTION MASS SPECTROMETRY

FIG. 2.

47

Collision-induced dissociation mass spectra of (A) ergosterol, (B) lanosterol, (C) cholesterol, and (D) squalene.

not accompanied by further fragmentation of the molecules studied. Previous work on triglycerides have demonstrated that neutral, nonpolar molecules can be ionized and are readily analyzed by APcI (14, 15), yielding mainly protonated molecular ions ([M ⫹ H] ⫹) and some diglyceride cations ([M ⫺ RCOO] ⫹) (14, 16). The minimal fragmentation of the acyltriglycerides can be of diagnostic value for identifying the positional association of the fatty acyl moiety with the glycerol backbone (14).

Collision-induced fragmentation of the four compounds produced the product ion spectra depicted in Figs. 2A–2D. As expected, similarities in the fragment patterns of all the compounds were detected. The molecular ion of ergosterol yielded two major product ions corresponding to m/z 69 and 57, with minor fragments at m/z 41, 81, 107, and 147 (Fig. 2A). The ions with m/z 69, 81, 95, and 109 were the major product ions present in the lanosterol, cholesterol, and squalene spectra (Figs. 2B–2D). The other fragments observed

48

TOH, PRIOR, AND VAN DER MERWE

FIG. 3. Analyte specificity of the ergosterol- and squalene-detecting channels. Plots of the intensity of the parent/product ion combination vs time showing that injection of squalene gave no response in the ergosterol-detecting channel (top right) and the injection of ergosterol gave no response in the squalene-detecting channel (bottom left) at the corresponding elution times.

were at m/z 57, 121, 135, and 147/149. It is of interest to compare the ergosterol mass spectrum obtained here with that obtained by GC/MS (8). Ionization and fragmentation by EI also yielded a m/z 69 fragment ion as the base peak, with the fragments at m/z 55, 81, and 363 also at relatively high abundance. A weak M •⫹ (m/z ⫽ 396) and low abundance [M ⫺ H 2O] •⫹ ion at m/z 378 were also detected, in contrast to the ApcI spectrum (8). Method Specificity For the quantification process, the method’s specificity for ergosterol and potential interference by structurally related compounds were checked by testing for signal crossover by lanosterol, cholesterol, and squalene. The tandem mass analyzers of the triplequadrupole instrument were set so that the first analyzer detected the precursor ions of ergosterol (m/z 380), lanosterol (m/z 410), cholesterol (m/z 370), or squalene (m/z 412). Since the fragment at m/z 69 was prominent in both the ergosterol and squalene fragmentation spectra, the second analyzer was, therefore, set to detect this product ion for both compounds. No squalene could be detected in the membrane extracts that were analyzed (data not shown). Although the same product ion was selected for the detection of both ergosterol and squalene, the use of multiple reaction monitoring as described above enabled the discrimination between the two compounds, as illustrated by the plots of the intensity of the specific precursor/product ion combination against time for the ergosterol- and squalene-detecting channels (Fig. 3). Injection of ergos-

terol alone produced a signal only in the ergosteroldetecting channel, whereas injection of squalene alone produced a response in the squalene-detecting channel only, with no response in the ergosterol channel. This is indicated by the lack of a signal at the corresponding elution times. The injection of lanosterol and cholesterol standards likewise did not give any response in the ergosterol-detecting channel (Fig. 4), confirming that APcI/MS/MS is able to provide the specificity required for differentiating between structurally related sterols. The robustness makes the method very suitable for the analysis of saponified extracts of whole cells (discussed in the next section) which in addition to ergosterol also contain precursors of ergosterol and other neutral lipids. Although the final membrane extract was usually dissolved in 2 ml ethanol prior to analysis, a comparison was made by dissolving the final extract in 2 ml hexane. Standards were similarly prepared and analyzed as described above. The results were the same as for analysis of extracts in ethanol, proving that the method is robust enough to accommodate the use of either ethanol or hexane for the solvent of the final preparation (results not shown). Linearity, Range, and Reproducibility A linear response with a coefficient of determination of 0.9992 was obtained with saponified ergosterol standards ranging from 0.15 to 5 ␮g/ml, corresponding to 7.56 –252 pmol ergosterol per 20 ␮l injection (Fig. 5). The reproducibility of the method at two different ergosterol concentrations is presented in Table 1. Six

ERGOSTEROL QUANTIFICATION BY DIRECT-INJECTION MASS SPECTROMETRY

49

FIG. 4. Analyte specificity of the ergosterol-detecting channel. Plots of the intensity of the parent/product ion combination vs time demonstrating that injection of the structural analogues lanosterol (50 ␮g/ml) and cholesterol (50 ␮g/ml) gave no response in the ergosterol channel (middle and bottom), indicating the specificity of the channel for ergosterol (top).

separate injections of the ergosterol standards gave a coefficient of variation of 3.33 and 1.98% for the 0.95 and the 4.25 ␮g/ml standard, respectively. Ergosterol extracted from purified plasma membranes of S. cerevisiae was analyzed by the method developed above and the results of three separate experiments are presented in Table 1. The ergosterol determination of the three independently extracted membrane samples gave a standard deviation of 41.71 ␮g/ml and a coefficient of variation of 11.18%. Since the analysis of the samples was carried out on different days with a fresh standard curve for each determination, the standard deviation and coefficient of variation are indicative of

the reproducibility and reliability of the method. The membrane ergosterol content of our strain determined by this method compares favorably with the value previously reported by Zinser et al. (1), who used gas chromatography and HPLC for ergosterol quantification. The coupling of GC separation does allow for lower detection limits, which, depending on the derivatization method, can be as low as 10 –20 pg injected ergosterol (8). LC/MS/MS can likewise offer low (ng/ml) detection limits with a high degree of specificity (23, 24). However, for the purpose of quantifying ergosterol in plasma membrane or whole cell extracts, the detection limit does not need to be lower than that achieved with this method. As an additional test of the robustness of direct injection APcI, unsaponified membrane lipids extracted by chloroform:methanol (2:1), were resolubilized in 2 ml ethanol and analyzed as described above. Due to the lower extraction efficiency of this procedure, a value of 242.7 ␮g ergosterol/mg protein for the same yeast strain was obtained. This value, nonetheless, lies within a similar order of magnitude to that reported in Table 1. The injection of the crude membrane lipid extract appears to be tolerated by the system and confirms the robustness of the method because the presence of other lipids (glycolipids, neutral lipids, and especially phospholipids) in the sample did not result in a spuriously high ergosterol value. CONCLUSION

FIG. 5. Standard curve relating signal strength of the m/z 69 product ion from ergosterol/squalene to ergosterol concentration.

The method described here was developed to quantify ergosterol without the need for prior fractionation either with GC or HPLC. This would be useful for

50

TOH, PRIOR, AND VAN DER MERWE TABLE 1

Reproducibility of Ergosterol Quantitation by Direct Injection APcI Reproducibility of ergosterol standard injections

a

Parameters

0.95 ␮g/ml

4.25 ␮g/ml

PM a ergosterol determination (␮g ergosterol/mg protein)

Mean ⫾ mean deviation No. of replicates Standard deviation Coefficient of variation

0.92 ⫾ 0.025 6 0.031 33.3%

4.25 ⫾ 0.07 6 0.084 1.98%

373.14 ⫾ 41.7 4 41.7 11.18%

Plasma membrane.

laboratories that do not have access to an analytical facility that is dedicated to routine ergosterol analysis. Therefore, the need to purchase specialized analytical columns for separation, associated instruments, and the optimization of analytical procedure for infrequent ergosterol analyses would make it an unjustifiable expense. To date, the use of direct injection APcI/MS/MS for ergosterol quantitation has not, to our knowledge, been reported. The successful application of APcI/ MS/MS for direct quantitation of thiourea has, however, been reported (10), with detection limits down to 1 ppb and a regression coefficient of 0.999931 for the calibration curve. The possibility of direct sample introduction into the APcI source has reduced the complexity and time associated with sample preparation and analysis. The use of MS/MS provided the specificity required for quantifying ergosterol without interference from structurally related sterols. This makes the method highly specific for ergosterol and obviates the need for the coupling of chromatographic separation with mass spectrometry or the need to derivatize the sterols. Although saponified samples were used here, this method works equally well with unsaponified samples containing free ergosterol. The data presented here demonstrate that direct injection APcI can be used for the quantification of plasma membrane ergosterol and is a reliable method with a high degree of specificity. ACKNOWLEDGMENTS We thank Professor P. Swart and Mr. G. Scheepers for helpful discussions and for providing the chemical structures. Support from the National Research Foundation to B.A.P. is gratefully acknowledged.

REFERENCES 1. Zinser, E., Paltauf, F., and Daum, G. (1993) Sterol composition of yeast organelle membranes and subcellular distribution of enzymes involved in sterol metabolism. J. Bacteriol. 175, 2853– 2858. 2. Parks, L. W., McLean-Bowen, C., Taylor, G. R., and Hough, S. (1978) Sterols in yeast subcellular fractions. Lipids 13, 730 –735.

3. Nurminen, T., Konttinen, K., and Suomalainen, H. (1975) Neutral lipids in the cells and cell envelope fractions of aerobic baker’s yeast and anaerobic brewer’s yeast. Chem. Physiol. Lipids 14, 15–32. 4. Shaw, W. H. C., and Jefferies, J. P. (1953) The determination of ergosterol in yeast. Analyst 78, 514 –519. 5. Arnezeder, C., Koliander, W., and Hampel, W. A. (1989) Rapid determination of ergosterol in yeast cells. Anal. Chim. Acta 225, 129 –136. 6. Rodriguez, R. J., and Parks, L. W. (1985) High performance liquid chromatography of sterols: Yeast sterols. Methods Enzymol. 111, 37–51. 7. Saraf, A., Larsson, L., Burge, H., and Milton, D. (1997) Quantification of ergosterol and 3-hydroxy fatty acids in settled house dust by gas chromatography-mass spectrometry: Comparison with fungal culture and determination of endotoxin by a Limulus amebocyte lysate assay. Appl. Environ. Microbiol. 63, 2554 – 2559. 8. Axelsson, B. O., Saraf, A., and Larsson, L. (1995) Determination of ergosterol in organic dust by gas chromatography-mass spectrometry. J. Chromatogr. B Biomed. Appl. 666, 77– 84. 9. McEwan, C. N., and Larson, B. (1998) in Mass Spectrometry of Biological Materials. (Larsen, B., and McEwan, C. N., Eds.), pp. 1–27, Dekker, New York. 10. Rafaelli, A. (1997) in Selected Topics and Mass Spectrometry in the Biomolecular Sciences (Caprioli, R. M., Malorni, A., and Sindona, G., Eds.), pp. 17–31, Kluwer Academic, Dordrecht, The Netherlands. 11. Siuzdak, G. (1996) Mass Spectrometry for Biotechnology. pp. 8-9, 14-18, Academic Press, San Diego. 12. Lee, E. D., Bruins, A. P., and Henion, J. D. (1986) Liquid chromatography/mass spectrometry. Anal. Chem. 58, 1451A–1460A. 13. Chapman, J. R. (1995) Practical Organic Mass Spectrometry. A Guide for Chemical and Biochemical Analysis. pp. 239 –243, Wiley, New York. 14. Laakso, P., and Voutilainen, P. (1996) Analysis of triacylglycerols by silver-ion high-performance liquid chromatography–atmospheric pressure chemical ionisation mass spectrometry. Lipids 31, 1311–1322. 15. Byrdwell, W. C., Emken, E. A., Neff, W. E., and Adlof, R. O. (1996) Quantitative analysis of triglycerides using atmospheric pressure chemical ionisation mass spectrometry. Lipids 31, 919 –935. 16. Byrdwell, W. C., and Emken, E. A. (1995) Analysis of triglycerides using atmospheric pressure chemical ionisation mass spectrometry. Lipids 30, 173–175.

ERGOSTEROL QUANTIFICATION BY DIRECT-INJECTION MASS SPECTROMETRY 17. Serrano, R. (1988) H ⫹-ATPase from plasma membranes of Saccharomyces cerevisiae and Avena sativa roots. Purification and reconstitution. Methods Enzymol. 157, 533–544. 18. Gessner, M. O., Bauchrowitz, M. A., and Escautier, M. (1991) Extraction and quantification of ergosterol as a measure of fungal biomass in leaf litter. Microb. Ecol. 22, 285–291. 19. Duncan, M. W. (1997) in Selected Topics and Mass Spectrometry in the Biomolecular Sciences (Caprioli, R. M., Malorni, A., and Sindona, G., Eds.), pp. 103–109, Kluwer Academic, Dordrecht, The Netherlands. 20. Green, J. M. (1996) A practical guide to analytical method validation. Anal. Chem. 68, 305A–309A. 21. Budzikiewicz, H., Djerassi, C., and Williams, D. (1967) Mass

51

Spectrometry of Organic Compounds, p. 113, Holden-Day, San Francisco. 22. Nes, W. R. (1985) A comparison of methods for the identification of sterols. Methods Enzymol. 111, 3–37. 23. Bo, L. D., Mazzucchelli, P., and Marzo, A. (1999) Highly sensitive bioassay of lidocaine in human plasma by high performance liquid chromatography–tandem mass spectrometry. J. Chromatogr. A 854, 3–11. 24. Fraser, I. J., Clare, R. A., and Pleasance, S. (1995) Determination of the acylcoenzyme A cholesterol acyltransferase inhibitor 447C88 in plasma using gas chromatography–mass spectrometry and liquid chromatography atmospheric pressure chemical ionisation tandem mass spectrometry. J. Chromatogr. B 663, 275–287.